Mine multi-dimensional fusion intelligent monitoring and early warning system method

Abstract

The invention relates to the technical field of coal mine dynamic disaster monitoring and early warning, and discloses a mine multi-dimensional fusion intelligent monitoring and early warning system and method. The system is a whole time-space integrated with twelve modules including an original data acquisition module, a mining condition analysis module, a data transmission module, data early-stage preprocessing, expert analysis and demonstration, optimal parameter delineation, basic theory calculation, numerical simulation analysis, laboratory verification, machine learning and neural network optimization, coupling model construction and field application detection. The invention discloses a multi-scale and multi-parameter coupling monitoring and early warning model and system. According to the mine multi-dimensional fusion intelligent monitoring and early warning system and method provided by the invention, a multi-level means is utilized, the limitation of a traditional single scheme or the subjectivity of single human interference is avoided, and the purpose of effectively preventing mine dynamic disasters can be achieved to a great extent; the method is of great significance to monitoring and early warning of coal rock gas hybrid dynamic disasters and safety production of mines.

Description

Technical Field

[0001] The present invention relates to the field of coal mine power disaster monitoring and early warning technology, and in particular to a multi-dimensional integrated intelligent monitoring and early warning system method for a mine. Background Art

[0002] my country is a major coal producer and consumer. The demand for coal continues to grow, and the depth and intensity of mining are constantly increasing, which leads to increasingly complex geological and mining conditions. The frequency and degree of harm of coal-rock gas compound dynamic disasters in mines are on the rise. In addition, the formation mechanism and evolution law of coal-rock gas compound dynamic disasters are not yet fully understood, and the existing prevention and control technologies have certain limitations, making it difficult to effectively respond to complex and changeable disaster situations.

[0003] In-depth research on coal-rock gas composite dynamic disasters will help reveal the internal mechanism of their occurrence and development, and provide a scientific basis for establishing more accurate prediction models and early warning systems, so as to take effective preventive measures in advance and reduce the possibility of disasters. This can not only protect the lives of coal miners, reduce casualties and property losses, but also improve the mining efficiency and safety of coal resources, and promote the sustainable development of the coal industry. At present, the main monitoring and early warning methods for mine dynamic disasters are microseismic method, drilling cuttings method, acoustic emission monitoring method, gas content assessment method, CT scanning technology, etc. They can achieve good monitoring effects to a certain extent, but for mines where disasters are complicated, a single early warning method is no longer sufficient to support the front-line needs of mines. Of course, there are also some researchers who combine two or a few monitoring technologies for monitoring, but there are still large inaccuracies and limitations.

[0004] In view of this, it is urgent to build a full-time, multi-scale, multi-parameter coupled monitoring and early warning model and system from the collection of original data, mining conditions analysis, data preprocessing, expert analysis, optimal parameter selection, monitoring system design, basic theoretical calculations, machine learning and neural network optimization, coupling model construction, numerical simulation analysis, laboratory verification, field application detection to early warning system development, a complete mine coal-rock gas composite dynamic disaster monitoring and early warning system "from head to toe, from original parameters to successful early warning" to effectively ensure the safety of mine front-line operations. Summary of the invention

[0005] The purpose of the present invention is to provide a multi-dimensional integrated intelligent monitoring and early warning system and implementation plan for mines to solve the technical problems existing in existing measures. This method intends to monitor the possibility of coal-rock gas compound dynamic disasters from multiple levels through multi-parameter fusion, and judge the hazard of dynamic disasters, and automatically issue early warnings based on the monitoring results, aiming to provide technical support and guarantee for safe production in coal mines.

[0006] To achieve the above purpose, the technical solution adopted by the present invention is as follows:

[0007] The multi-dimensional integrated intelligent monitoring and early warning system for mines proposed in the present invention includes twelve modules, namely, original data acquisition module, mining condition analysis module, data transmission module, data preprocessing module, expert analysis and demonstration, optimal parameter delineation module, basic theoretical calculation, numerical simulation analysis module, experimental verification module, machine learning and neural network optimization module, coupled monitoring and early warning model and system and field verification.

[0008] Preferably, the raw data collection should deploy multiple sensors and monitoring equipment to collect environmental data in the mine, such as gas concentration, gas pressure, drill cuttings data, electromagnetic parameters, microseismic data, temperature, humidity, etc., to ensure the accuracy and comprehensiveness of data collection, and calibrate the equipment regularly;

[0009] Preferably, the mining condition analysis requires a detailed analysis of the geological structure, mining method, operation intensity and other conditions of the mine, to clarify the distribution of the underground coal and rock layer structure, to clarify the regional scope of faults, folds and other areas prone to compound dynamic disasters that may occur during the mining process, to consider the impact of mining conditions on monitoring data, and to provide a basis for subsequent analysis;

[0010] Preferably, the data transmission is the key link between the data collection point and the data processing center. Select appropriate communication technology, such as wireless communication, wired network or optical fiber communication. Establish a stable data transmission network to ensure the stability and reliability of the data transmission network to reduce data loss and transmission delay. The terminals of the raw data collection module and the mining condition analysis module are connected to the data transmission module;

[0011] Preferably, the data pre-processing module is intended to perform pre-processing such as cleaning, denoising, and standardization on the collected raw data. In this process, data loss and information distortion must be avoided to ensure data quality; it is connected to the data transmission module terminal, and the processed data are all transmitted by the communication technology in the data transmission module;

[0012] Preferably, the expert analysis and demonstration refers to inviting industry experts to analyze the data processed by the data pre-processing module and provide professional opinions, but it is necessary to ensure that the opinions and suggestions of the experts are highly objective and popular, and can ensure the validity of the data analysis results and can be fully understood and applied by the system designer;

[0013] Preferably, the optimal parameter delineation refers to evaluating the influence of different parameters on mine safety through correlation analysis, principal component analysis (PCA) and other methods, retaining the relevant parameters that can best reflect and monitor coal-rock gas composite dynamic disasters in combination with the expert analysis and demonstration module, and then performing sensitivity analysis on key parameters to understand the influence of parameter changes on the performance of the early warning system, and finally using optimization algorithms (such as genetic algorithms, gradient descent methods, etc.) to determine key optimal parameters, such as microseismic, gas pressure, gas content, ground stress and other data, and then transmitting them to the corresponding database of the computer through another port of the data transmission module, such as microseismic data is a, gas content is b, etc.;

[0014] Preferably, the basic theoretical calculation needs to use professional theoretical calculation formulas (such as stiffness theory, impact tendency theory, "three criteria" theory, deformation instability theory, energy theory, strength theory, "rheological hypothesis", "fluid-solid coupling instability theory", "spherical shell instability hypothesis", "three factors" theory, etc.), and the corresponding database of the computer is called out in combination with the theoretical calculation formula and the above-mentioned computer database code. It is worth mentioning that the corresponding parameters in the above-mentioned theoretical calculation formula need to be replaced with the corresponding data code of the corresponding database;

[0015] Preferably, the numerical simulation analysis needs to design simulation details according to the above parameters, including different working conditions, parameter changes, etc., to comprehensively evaluate the safety status of the mine, such as using numerical simulation software such as Comsol, PFC, FLAC3D to simulate a single power disaster, and then using the secondary development solution to analyze its composite power disaster, identify the potential risk points and safety thresholds of the mine, adjust the simulation parameters, and conduct a large number of corrections;

[0016] Preferably, the laboratory verification can design a gas-solid coupling experimental system according to the coal-rock gas composite dynamic disaster, set the experimental conditions according to the relevant optimal parameters, study the potential inducing critical value of the composite dynamic disaster, and conduct multiple experiments. It should be noted that the variable control during the experiment must be guaranteed;

[0017] Preferably, the machine learning and neural network optimization are performed on the basis of the parameter revision of the data pre-processing module, expert analysis and demonstration, and the optimal parameter delineation module. First, the neural network architecture needs to be designed, including the number of layers, the number of neurons, the activation function, etc., and then the model performance is optimized using methods such as cross-validation and hyperparameter tuning, and then the use of integrated learning methods, such as random forests and gradient boosting machines, is considered to improve the generalization ability of the model; further, the performance of the model is evaluated using a test set, focusing on indicators such as accuracy, recall, and F1 scores, and then the results obtained from the above-mentioned large amount of basic theoretical calculations, numerical simulation analysis, and laboratory verification are placed in the machine learning and neural network optimization module for training and optimization, and finally the induced critical values ​​and basic models under different conditions are obtained, such as theoretical models, simulation models, and experimental models;

[0018] Preferably, the coupling model construction refers to the combination of the theoretical model, the simulation model and the experimental model, and the typical dynamic disaster mine parameters are "fed" into the computer database respectively, and a large number of verifications are carried out to judge the differences between the three. Experts are invited to analyze again, and the three major models are optimized to extract the essence and remove the dross, and finally a relatively complete full-time, multi-scale, multi-parameter coupling monitoring and early warning model and system is obtained.

[0019] On the other hand, the present invention also provides a specific design step of a multi-dimensional integrated intelligent monitoring and early warning solution for a mine:

[0020] S1. Define the monitoring scope and area, and reasonably arrange point, surface, static and dynamic multi-level monitoring equipment, such as microseismic sensors, gas pressure sensors and other monitoring equipment, so that the entire monitoring mine is fully covered and data within the complete range is collected;

[0021] S2. Analyze the specific laws of mining conditions in the area, clarify the geological structure, mining methods, operation intensity and other conditions of the mine, and clarify the distribution of coal and rock structure underground;

[0022] S3. Invite experts and scholars in the industry to evaluate and screen effective parameters and enter them into the monitoring database, and then determine the number of monitoring points based on the effective parameters of the mine, that is, the number of collection layout points 3 (such as electromagnetic radiation pulses, number of microseisms, gas pressure, gas concentration, etc.);

[0023] S4. Establish a stable data transmission network to ensure the stability and reliability of the data transmission network to reduce data loss and transmission delay;

[0024] S5. Perform pre-processing of the collected raw data, such as cleaning, denoising, and standardization;

[0025] S6. Invite industry experts to analyze the data after pre-processing and provide professional opinions to ensure the selection of optimal parameter values;

[0026] S7. Evaluate the influence of different parameters on mine safety through correlation analysis, principal component analysis (PCA) and other methods, retain the relevant parameters that can best reflect and monitor coal-rock gas compound dynamic disasters in combination with expert analysis and demonstration, and then perform sensitivity analysis on key parameters to determine the key optimal parameters, and then transmit them to the computer database through another port of the data transmission module, such as microseismic data is a, gas content is b, etc.;

[0027] S8, combining the theoretical calculation formula and the above-mentioned computer database code to call out the computer database, and bring the parameters into the theoretical calculation formula to perform a large number of calculation predictions;

[0028] S9. Design simulation details based on the above parameters, including different working conditions, parameter changes, etc., to comprehensively evaluate the safety status of the mine, identify potential risk points and safety thresholds of the mine through a large number of simulations, adjust simulation parameters, and conduct a large number of analyses;

[0029] S10. Design a reasonable gas-solid coupling experimental system for coal-rock gas composite dynamic disasters, set experimental conditions according to relevant optimal parameters, and conduct multiple experiments;

[0030] S11. Place the results obtained from the above-mentioned large number of theoretical calculation modules, numerical simulation analysis modules and experimental verification modules in the machine learning and neural network optimization modules for training and optimization, and obtain the induced critical values ​​and basic models under different circumstances, such as theoretical models, simulation models and experimental models. Finally, deploy the trained three models into the monitoring and early warning system;

[0031] S12. Combining the theoretical model, simulation model and experimental model, respectively "feeding" the typical dynamic disaster mine parameters into the optimization database, conducting a large number of verifications, judging the differences between the three, and again asking experts to analyze and optimize the three models, extracting the essence and removing the dross, and finally obtaining a relatively complete full-time, multi-scale, multi-parameter coupling monitoring and early warning model and system;

[0032] S13. Select multiple mines to collect data, input the data into the coupled monitoring and early warning model and system, conduct on-site verification, and check the reliability and accuracy of the model.

[0033] Compared with the related technologies, the multi-dimensional integrated intelligent monitoring and early warning system method for mines proposed in the present invention has the following beneficial effects:

[0034] The present invention proposes a multi-dimensional integrated intelligent monitoring and early warning system method for mines:

[0035] 1. The multi-dimensional integrated intelligent monitoring and early warning scheme for mines proposed in the present invention integrates multi-level means such as theoretical calculation, numerical analysis, experimental verification, big data optimization and expert professional analysis, avoiding the limitations of traditional single schemes or the subjectivity of single human intervention, and can achieve the purpose of effectively preventing mine power disasters to a great extent. It is of great significance to reduce the monitoring and early warning of coal-rock gas composite power disasters and the safe production of mines;

[0036] 2. In the multi-dimensional integrated intelligent monitoring and early warning scheme for mines proposed in the present invention, three large models (theoretical model, simulation model and experimental model) are subjected to a large number of optimization processes of deep learning and neural networks, and the three are compared and analyzed. The resulting full-time, multi-scale, multi-parameter coupled monitoring and early warning model and system can cover all areas and time periods of the mine, ensuring that there are no blind spots and providing continuous monitoring data;

[0037] 3. The multi-dimensional integrated intelligent monitoring and early warning solution for mines proposed in the present invention integrates multiple data sources and monitoring technologies, improves the richness and reliability of data, and helps to more accurately identify and predict complex phenomena in mines. At the same time, the data transmission module ensures real-time transmission of data, enabling the monitoring and early warning system to respond quickly and issue early warnings in a timely manner. BRIEF DESCRIPTION OF THE DRAWINGS

[0038] Figure 1 This is a diagram illustrating the specific steps of the multi-dimensional integrated intelligent monitoring and early warning system method for mines proposed by the present invention;

[0039] Figure 2 This is a specific flow chart of the multi-dimensional integrated intelligent monitoring and early warning system method for mines proposed in the present invention.

[0040] Numbers in the figure: 1. Data acquisition module; 2. Mining condition analysis module; 3. Collection layout points; 4. Expert analysis and demonstration; 5. Monitoring equipment; 6. Data transmission module; 7. Data preprocessing module; 8. Optimal parameter delineation module; 9. Basic theoretical calculation; 10. Theoretical calculation formula; 11. Numerical simulation analysis module; 12. Laboratory verification; 13. Machine learning and neural network optimization module; 14. Theoretical model; 15. Simulation model; 16. Experimental model; 17. Optimization database; 18. Coupled monitoring and early warning model and system; 19. Field verification. DETAILED DESCRIPTION

[0041] The present invention will be further described below in conjunction with the accompanying drawings and implementation modes.

[0042] Please refer to Figure 1-2 ,in, Figure 1A schematic diagram of a preferred embodiment of the multi-dimensional integrated intelligent monitoring and early warning system method for mines proposed by the present invention; Figure 2 for Figure 1 An enlarged schematic diagram of part A is shown.

[0043] The present invention proposes a multi-dimensional integrated intelligent monitoring and early warning system for mines, comprising twelve modules, namely, an original data acquisition module 1, a mining condition analysis module 2, a data transmission module 6, a data pre-processing module 7, an expert analysis and demonstration module 4, an optimal parameter delineation module 8, a basic theoretical calculation 9, a numerical simulation analysis module 11, an experimental verification module 12, a machine learning and neural network optimization module 13, a coupled monitoring and early warning model and system 18 and a field verification 19.

[0044] The raw data collection 1 should deploy multiple sensors and monitoring equipment 5 to collect environmental data in the mine, such as gas concentration, gas pressure, drill cuttings data, electromagnetic parameters, microseismic data, temperature, humidity, etc., to ensure the accuracy and comprehensiveness of data collection, and calibrate the equipment regularly;

[0045] The mining condition analysis requires a detailed analysis of the geological structure, mining method, operation intensity and other conditions of the mine, clarifying the distribution of coal and rock structure underground, clarifying the regional scope of faults, folds and other complex dynamic disasters that may occur during the mining process, considering the impact of mining conditions on monitoring data, and providing a basis for subsequent analysis;

[0046] The data transmission is the key link between the data collection point and the data processing center. Choose the appropriate communication technology, such as wireless communication, wired network or optical fiber communication. Establish a stable data transmission network to ensure the stability and reliability of the data transmission network to reduce data loss and transmission delay. The terminals of the raw data collection module 1 and the mining condition analysis module 2 are connected to the data transmission module 6;

[0047] The data preprocessing is to clean, denoise, standardize and other preprocessing of the collected raw data. In this process, data loss and information distortion should be avoided to ensure data quality; it is connected to the data transmission module 6 terminal, and the processed data are transmitted by the communication technology in the data transmission module 6;

[0048] The expert analysis and demonstration 4 refers to inviting industry experts to analyze the data after the pre-processing of the data and provide professional opinions, but it is necessary to ensure that the opinions and suggestions of the experts are highly objective and popular, and can ensure the validity of the data analysis results and can be fully understood and applied by the system designer;

[0049] The optimal parameter delineation refers to evaluating the influence of different parameters on mine safety through correlation analysis, principal component analysis (PCA) and other methods, and retaining the relevant parameters that can best reflect and monitor coal-rock gas compound dynamic disasters in combination with expert analysis and demonstration, and then conducting sensitivity analysis on key parameters to understand the influence of parameter changes on the performance of the early warning system, and finally using optimization algorithms (such as genetic algorithms, gradient descent methods, etc.) to determine key optimal parameters, such as microseismic data, gas pressure, gas content, ground stress and other data, and then transmitting them to a computer database, such as microseismic data is a, gas content is b, etc.;

[0050] The basic theoretical calculation 9 needs to use professional theoretical calculation formula 10 (such as stiffness theory, impact tendency theory, "three criteria" theory, deformation instability theory, energy theory, strength theory, "rheological hypothesis", "fluid-solid coupling instability theory", "spherical shell instability hypothesis", "three factors" theory, etc.), and the corresponding database of the computer is called out in combination with the theoretical calculation formula 10 and the above-mentioned computer database code. It is worth mentioning that the corresponding parameters in the above-mentioned theoretical calculation formula 10 need to be replaced with the corresponding data code of the corresponding computer database;

[0051] The numerical simulation analysis needs to design simulation details according to the above parameters, including different working conditions, parameter changes, etc., to comprehensively evaluate the safety status of the mine, such as using numerical simulation software such as Comsol, PFC, FLAC3D to simulate a single power disaster, and then using the secondary development solution to analyze its compound power disasters, identify the potential risk points and safety thresholds of the mine, adjust the simulation parameters, and conduct a large number of corrections;

[0052] The laboratory verification can design a gas-solid coupling experimental system according to the coal-rock gas compound dynamic disaster, set the experimental conditions according to the relevant optimal parameters, study the potential inducing critical value of the compound dynamic disaster, and conduct multiple experiments. It should be noted that the variable control during the experiment must be guaranteed;

[0053] The machine learning and neural network optimization are carried out on the basis of the parameter revision of the data pre-processing module 7, the expert analysis and demonstration 4 and the optimal parameter delineation module 8. First, the neural network architecture needs to be designed, including the number of layers, the number of neurons, the activation function, etc., and then the model performance is optimized by using methods such as cross-validation and hyperparameter tuning, and then the use of integrated learning methods, such as random forests and gradient boosting machines, is considered to improve the generalization ability of the model; further, the performance of the model is evaluated using a test set, focusing on indicators such as accuracy, recall rate, and F1 score, and then the results obtained by the above-mentioned large amount of basic theoretical calculations 9, numerical simulation analysis and laboratory verification are placed in the machine learning and neural network optimization module 13 for training optimization, and finally the induced critical value and basic model under different conditions are obtained, such as the theoretical model 14, the simulation model 15 and the experimental model 16;

[0054] The construction of the coupling model refers to the combination of the theoretical model 14, the simulation model 15 and the experimental model 16. The typical dynamic disaster mine parameters are "fed" into the optimization database 17 respectively, and a large number of verifications are carried out to judge the differences between the three. Experts are invited to analyze again, optimize the three models, extract the essence and remove the dross, and finally obtain a relatively complete full-time, multi-scale, multi-parameter coupling monitoring and early warning model and system 18.

[0055] In addition, the present invention also proposes a multi-dimensional integrated intelligent monitoring and early warning method for mines, which specifically includes the following steps:

[0056] S1. Define the monitoring scope and area, and reasonably arrange point, surface, static and dynamic multi-level monitoring equipment, such as microseismic sensors, gas pressure sensors and other monitoring equipment, so that the entire monitoring mine is fully covered and data within the complete range is collected;

[0057] S2. Analyze the specific laws of mining conditions in the area, clarify the geological structure, mining methods, operation intensity and other conditions of the mine, and clarify the distribution of coal and rock structure underground;

[0058] S3. Invite experts and scholars in the industry to evaluate and screen effective parameters and enter them into the monitoring database, and then determine the number of monitoring points based on the effective parameters of the mine, that is, the number of collection layout points 3 (such as electromagnetic radiation pulses, number of microseisms, gas pressure, gas concentration, etc.);

[0059] S4. Establish a stable data transmission network to ensure the stability and reliability of the data transmission network to reduce data loss and transmission delay;

[0060] S5. Perform pre-processing of the collected raw data, such as cleaning, denoising, and standardization;

[0061] S6. Invite industry experts to analyze the data after pre-processing and provide professional opinions to ensure the selection of optimal parameter values;

[0062] S7, evaluate the influence of different parameters on mine safety through correlation analysis, principal component analysis (PCA) and other methods, retain the relevant parameters that can best reflect and monitor coal-rock gas compound dynamic disasters in combination with expert analysis and demonstration, and then perform sensitivity analysis on key parameters to determine the key optimal parameters, and then transmit them to the computer database through another port of the data transmission module 6, such as microseismic data is a, gas content is b, etc.;

[0063] S8, combining the theoretical calculation formula and the above-mentioned computer database code to call out the computer database, and bring the parameters into the theoretical calculation formula to perform a large number of calculation predictions;

[0064] S9. Design simulation details based on the above parameters, including different working conditions, parameter changes, etc., to comprehensively evaluate the safety status of the mine, identify potential risk points and safety thresholds of the mine through a large number of simulations, adjust simulation parameters, and conduct a large number of analyses;

[0065] S10. Design a reasonable gas-solid coupling experimental system for coal-rock gas composite dynamic disasters, set experimental conditions according to relevant optimal parameters, and conduct multiple experiments;

[0066] S11, placing the results obtained by the large number of theoretical calculation modules 10, numerical simulation analysis modules 11 and experimental verification modules 12 in the machine learning and neural network optimization module 13, performing training optimization, and obtaining the induced critical values ​​and basic models under different conditions, such as the theoretical model 14, the simulation model 15 and the experimental model 16, and finally deploying the trained three major models into the monitoring and early warning system;

[0067] S12, combining the theoretical model 14, the simulation model 15 and the experimental model 16, respectively "feeding" the typical dynamic disaster mine parameters to the optimization database 17, conducting a large number of verifications, judging the differences between the three, and again asking experts to analyze and optimize the three models, taking the essence and removing the dross, and finally obtaining a relatively complete full-time, multi-scale, multi-parameter coupling monitoring and early warning model and system 18;

[0068] S13, select multiple mines to collect data, input the data into the coupled monitoring and early warning model and system 18, conduct on-site verification 19, and verify the reliability and accuracy of the model.

[0069] Compared with related technologies, the multi-dimensional integrated intelligent monitoring and early warning system method for mines proposed in this invention has significant advancements:

[0070] First, this method integrates multiple means such as theoretical calculation, numerical analysis, experimental verification, big data optimization and expert professional analysis, overcoming the limitations of traditional single solutions and the subjectivity of human intervention, effectively preventing mine power disasters, and is of great significance to reducing coal-rock gas compound power disasters and ensuring mine safety production;

[0071] Secondly, by optimizing the processing of three large models through deep learning and neural network, and conducting comparative analysis, the resulting full-time, multi-scale, multi-parameter coupled monitoring and early warning model and system can fully cover all areas and time periods of the mine, without blind spots, and provide continuous monitoring data;

[0072] Finally, the integration of multiple data sources and monitoring technologies improves the richness and reliability of data, helps to more accurately identify and predict complex phenomena in mines, and the data transmission module ensures real-time transmission of data, enabling the monitoring and early warning system to respond quickly and issue warnings in a timely manner.

[0073] The above descriptions are merely embodiments of the present invention and are not intended to limit the patent scope of the present invention. Any equivalent structure or equivalent process transformation made using the contents of the present invention specification and drawings, or directly or indirectly applied in other related technical fields, are also included in the patent protection scope of the present invention.

Claims

1. A multi-dimensional integrated intelligent monitoring and early warning system method for mines, characterized in that: The steps include: S1. Define the monitoring scope and area, and reasonably arrange point, surface, static and dynamic multi-level monitoring equipment, such as microseismic sensors and gas pressure sensors, so that the entire monitoring mine is fully covered and data within the complete range is collected; S2. Analyze the specific laws of mining conditions in the area, clarify the geological structure, mining methods, and operating intensity conditions of the mine, and clarify the distribution of coal and rock structures underground; S3. Invite experts and scholars in the industry to evaluate and screen effective parameters and enter them into the monitoring database, and then determine the number of monitoring points based on the effective parameters of the mine, that is, the number of collection points such as electromagnetic radiation pulses, number of microseisms, gas pressure, and gas concentration; S4. Establish a stable data transmission network to ensure the stability and reliability of the data transmission network to reduce data loss and transmission delay; S5. Clean, denoise and standardize the collected raw data for preliminary data preprocessing; S6. Invite industry experts to analyze the data after pre-processing and provide professional opinions to ensure the selection of optimal parameter values; S7. Evaluate the influence of different parameters on mine safety through correlation analysis and principal component analysis, retain the relevant parameters that can best reflect and monitor coal-rock gas compound dynamic disasters in combination with expert analysis and demonstration, and then perform sensitivity analysis on key parameters to determine the key optimal parameters, and then transmit them to the computer database through another port of the data transmission module, such as microseismic data is a and gas content is b; S8, combining the theoretical calculation formula and the above-mentioned computer database code to call out the computer database accordingly, and bring the parameters into the theoretical calculation formula to perform a large number of calculation predictions; S9. Design simulation details based on the above parameters, including different working conditions and parameter changes, to comprehensively evaluate the safety status of the mine, identify potential risk points and safety thresholds of the mine through a large number of simulations, adjust simulation parameters, and conduct a large number of analyses; S10. Design a reasonable gas-solid coupling experimental system for coal-rock gas composite dynamic disasters, set experimental conditions according to relevant optimal parameters, and conduct multiple experiments; S11. Place the results obtained from the above-mentioned large number of theoretical calculation modules, numerical simulation analysis modules and experimental verification modules in the machine learning and neural network optimization modules for training and optimization, and obtain the induced critical values ​​and basic models under different circumstances, such as theoretical models, simulation models and experimental models. Finally, deploy the trained three models into the monitoring and early warning system; S12. Combining the theoretical model, simulation model and experimental model, respectively "feeding" the typical dynamic disaster mine parameters into the optimization database, conducting a large number of verifications, judging the differences between the three, and again asking experts to analyze and optimize the three models, extracting the essence and removing the dross, and finally obtaining a relatively complete full-time, multi-scale, multi-parameter coupling monitoring and early warning model and system; S13, select multiple mines to collect data, input the data into the coupled monitoring and early warning model and system 18, conduct on-site verification, and test the reliability and accuracy of the model.

2. The method for intelligent monitoring and early warning system for mines with multi-dimensional integration according to claim 1 is characterized in that: The expert analysis and demonstration refers to inviting industry experts to analyze the data after the preliminary pre-processing of the data and provide professional opinions, but it is necessary to ensure that the opinions and suggestions of the experts are highly objective and popular, and can ensure the validity of the data analysis results and can be fully understood and applied by the system designers.

3. The method for intelligent monitoring and early warning system of mine multi-dimensional fusion according to claim 1 is characterized in that: The optimal parameter delineation refers to evaluating the impact of different parameters on mine safety through correlation analysis and principal component analysis, retaining the relevant parameters that can best reflect and monitor coal-rock gas complex dynamic disasters in combination with expert analysis and demonstration, and then conducting sensitivity analysis on key parameters to understand the impact of parameter changes on the performance of the early warning system. Finally, an optimization algorithm is used to determine key optimal parameters, such as microseismic, gas pressure, gas content, and ground stress data, which are then transmitted to a computer database, such as microseismic data is a and gas content is b.

4. The method for intelligent monitoring and early warning system of mine multi-dimensional fusion according to claim 1 is characterized in that: The basic theoretical calculations need to use professional theoretical calculation formulas, such as stiffness theory, impact tendency theory, "three criteria" theory, deformation instability theory, energy theory, strength theory, "rheological hypothesis", "fluid-solid coupling instability theory", "spherical shell instability hypothesis", and "three factors" theory. The corresponding database of the computer is called out in combination with the theoretical calculation formulas and the above-mentioned computer database codes. It is worth mentioning that the corresponding parameters in the above-mentioned theoretical calculation formulas need to be replaced with the corresponding data codes of the corresponding computer database.

5. The method for intelligent monitoring and early warning system of mine multi-dimensional fusion according to claim 1 is characterized in that: The numerical simulation analysis needs to design simulation details according to the above parameters, including different working conditions and parameter changes, in order to comprehensively evaluate the safety status of the mine. For example, Comsol, PFC, and FLAC3D numerical simulation software are used to simulate a single power disaster, and then the secondary development solution is used to analyze its compound power disasters, identify the potential risk points and safety thresholds of the mine, adjust the simulation parameters, and conduct a large number of corrections.

6. The method for intelligent monitoring and early warning system of mine multi-dimensional fusion according to claim 1 is characterized in that: The laboratory verification can design a gas-solid coupling experimental system according to the coal-rock gas composite dynamic disaster, set the experimental conditions according to the relevant optimal parameters, study the potential inducing critical value of the composite dynamic disaster, and conduct multiple experiments. It should be noted that the variable control during the experiment must be guaranteed.

7. The method for intelligent monitoring and early warning system of mine multi-dimensional fusion according to claim 1 is characterized in that: The machine learning and neural network optimization are carried out on the basis of the parameter revision of the data preprocessing, expert analysis and demonstration, and optimal parameter delineation. First, the neural network architecture needs to be designed, including the number of layers, the number of neurons, and the activation function, and then the model performance is optimized using cross-validation and hyperparameter tuning methods, and then the use of integrated learning methods, such as random forests and gradient boosting machines, is considered to improve the generalization ability of the model; further, the performance of the model is evaluated using a test set, focusing on accuracy, recall, and F1 score indicators, and then the results obtained from the above-mentioned large amount of theoretical calculations, numerical simulation analysis, and experimental verification are placed in the machine learning and neural network optimization module for training and optimization, and finally the induced critical values ​​and basic models under different situations are obtained, such as theoretical models, simulation models, and experimental models.

8. The method for intelligent monitoring and early warning system of mine multi-dimensional fusion according to claim 1 is characterized in that: The construction of the coupling model refers to the combination of the theoretical model, simulation model and experimental model. The typical dynamic disaster mine parameters are "fed" into the optimization database respectively, and a large number of verifications are carried out to judge the differences between the three. Experts are invited to analyze again, optimize the three models, extract the essence and remove the dross, and finally obtain a relatively complete full-time, multi-scale, multi-parameter coupling monitoring and early warning model and system.

Images